10718
J. Agric. Food Chem. 2009, 57, 10718–10731 DOI:10.1021/jf902594m
Phytochemical Profiles and Health-Promoting Effects of Cool-Season Food Legumes As Influenced by Thermal Processing BAOJUN XU† AND SAM K. C. CHANG*,‡ †
Food Science and Technology Program, Beijing Normal University—Hong Kong Baptist University United International College, Zhuhai, Guangdong 519085, China, and ‡Department of Cereal and Food Sciences, North Dakota State University, Fargo, North Dakota 58108
The effects of four thermal processing methods (conventional boiling, conventional steaming, pressure boiling, and pressure steaming) on phytochemical profiles, antioxidant capacities, and antiproliferation properties of commonly consumed cool-season food legumes, including green pea, yellow pea, chickpea, and lentil, were investigated. Four groups of individual phenolic compounds, including phenolic acids, anthocyanins, and flavan-3-ols, as well as flavonols and flavones were quantified using HPLC, respectively. As compared to the original raw legumes, all processing methods caused significant (p < 0.05) reduction in total phenolic content, procyanidin content, total saponin content, phytic acid content, chemical antioxidant capacities in terms of ferric reducing antioxidant power and peroxyl radical scavenging capacity, and cellular antioxidant activity as well as antiproliferation capacities of cool-season food legumes. Different cooking methods have varied effects on reducing total phenolics, saponins, phytic acids, and individual phenolic compounds. For all cool-season food legumes, steaming appeared to be a better cooking method than boiling in retaining antioxidants and phenolic components, whereas boiling appeared to be effective in reducing saponin and phytic acid contents. In the case of lentil, all thermal processing methods (except conventional steaming) caused significant (p < 0.05) decreases in gallic, chlorogenic, p-coumaric, sinapic, subtotal benzoic, subtotal cinnamic acid, and total phenolic acid. All thermal processing methods caused significant (p < 0.05) decreases in (þ)-catechin and flavan-3-ols in each cool-season food legume. KEYWORDS: Cool-season food legumes; boiling; steaming; phenolic acid; flavan-3-ol; flavonol; flavone; anthocyanin; saponin; phytic acid; antioxidant; FRAP; PRSC; cellular antioxidant activity; antiproliferation; HPLC
INTRODUCTION
Food legumes are economical dietary sources of good-quality protein, carbohydrates, dietary fiber components, and a variety of minerals and vitamins. Food legumes contain several compounds that have been traditionally considered to be antinutrients, such as protease inhibitors, phytic acids, saponins, tannins, plant sterols, and isoflavones. However, more recent information suggests that most of these compounds may actually benefit the consumer’s health if used properly in the context of foods for disease prevention. Research suggests that regular dietary intake of food legumes can reduce the risk of developing nutritionrelated health problems including obesity, diabetes, heart diseases, and cancers (1). Therefore, food legumes are recommended as an excellent food choice with health-promoting benefits. The cool-season food legumes (CSFLs), including green pea, yellow pea, chickpea, and lentil, are traditionally low-input crops and are grown extensively in the farming system of the Indian subcontinent, the Mediterranean area, the Nile Valley, Central *Author to whom correspondence should be addressed [telephone (701) 231-7485; fax (701) 231-6536; e-mail
[email protected]].
pubs.acs.org/JAFC
Published on Web 10/29/2009
Europe, the Americas, and Australia (2). The U.S. production of CSFLs, mainly in the northwestern states, such as Washington, Idaho, Montana, and North Dakota, has increased significantly in recent years (3). The CSFLs have many nutritional qualities that make them attractive to food manufacturers. Legumes must be cooked before consumption. Although antioxidant properties and phenolic compounds of raw uncooked CSFLs have been reported (4, 5), how processing methods affect the phytochemicals and health-promoting activities, such as antioxidant activities and anticancer properties, has not been systematically studied. Our preliminary study showed that soaking, boiling, and steaming processes significantly affected the total phenolic contents, free radical scavenging activities, and oxygen radical absorbing capacities (ORAC) of CSFLs (6). To continue our study on thermal processing effects, the present study was undertaken to further investigate how thermal processing affected individual phenolic compounds (including phenolic acids and flavonoids), saponins, and phytic acids, as well as ferric reducing antioxidant power (FRAP) and peroxyl radical scavenging capacity (PRSC), cellular antioxidant activities (CAA), and antiproliferation activities of CSFLs.
© 2009 American Chemical Society
Article MATERIALS AND METHODS
Chemicals and Standards. Sixteen phenolic acids and three aldehydes, five flavan-3-ols [(þ)-catechin, (þ)-epicatechin, epigallocatechin, epicatechin-gallate, epigallatecatechin-gallate (EGCG)], six flavonols or flavones (myricetin, luteolin, quercetin, apigenin, kaempferol, quercetin-3rutinoside), soya saponin (contained a minimum of 80% saponin), phytic acid, sulfosalicylic acid, trifluoroacetic acid (TFA), Folin-Ciocalteu reagent, sodium carbonate, dimethyl sulfoxide (DMSO), 20 ,70 -dichlorofluorescin diacetate (DCFH-DA), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), 2,4,6-tri(2-pyridyl)-s-triazine (TPTZ), and 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). Kaempferol-3-O-glucoside, kaempferol-3-O-rutinoside, and quercetin-3O-glucoside were purchased from Extrasynthese S.A. (Genay, France). A mixture of six unimolar anthocyanin standards (3-O-β-glucosides of delphinidin, cyanidin, petunidin, pelargonidin, peonidin, and malvidin) was purchased from Polyphenols Laboratories (Sandnes, Norway). 2,20 Azobis(2-amidinopropane) dihydrochloride (AAPH) was purchased from Wako Chemicals USA (Richmond, VA). HPLC-grade solvents (methanol and acetonitrile, B&J Brand) and other analytical grade solvents used for extraction were purchased from VWR International (West Chester, PA). Human gastric adenocarcinoma cell line AGS, human colorectal adenocarcinoma cell line SW480, and human prostate carcinoma cell line DU145 were purchased from American Type Culture Collection (ATCC, Manassas, VA). Hanks balanced salt solution (HBSS) and 0.4% trypan blue stain solution were purchased from Cambrex Bio Science Walkersville, Inc. (Walkersville, MD). Phosphate-buffered saline (PBS), trypsin-EDTA solution, penicillin-streptomycin, fetal bovine serum (FBS), and cell culture media (Eagle’s MEM and F-12K) were purchased from Mediatech, Inc. (Herndon, VA). Legume Materials. The dried cool-season food legume (CSFL) seeds used in the current study were green pea (Pisum sativum L. cv. Stratus), supplied by Meridian Seed LLC (West Fargo, ND); yellow pea (P. sativum L. cv. Golden), supplied by Steve Marman Pulse USA (Bismark, ND); and chickpea (Cicer arietinum L. cv. Amits) and lentil (Lens culinaris cv. CDC Richlea), supplied by Agricare United (Ray, ND). Broken seeds, damaged seeds, and foreign materials were removed from the samples. Moisture content was determined by drying the samples in an air-circulated oven at 110 °C until a constant weight was obtained (7). All calculations for determination of phenolics and quantification of antioxidant activities are on a dry weight basis. Soaking and Hydration Ratio. The soaking procedures and determination method of hydration ratio in our earlier paper (6) were followed. Soaking time of CSFLs with desired hydration ratio was calculated by calibration through a quadratic fit equation of respective water adsorption curve as previously described (6). The soaked peas (with 100% hydration ratio) and lentils (with 50% hydration ratio) were drained and then boiled or steamed according to the methods described below. Boiling, Steaming, and Cooking Time. All thermal processes were performed according to the procedures published in our earlier paper (6). Briefly, conventional boiling and steaming treatments were conducted using a domestic atmospheric cooker and a domestic atmospheric steam cooker, respectively. Pressure boiling and steaming were conducted using an M-0512-H Mirro pressure cooker (Mirro Co., Manitowoc, WI), respectively. The cooking time was determined on the basis of a tactile method according to Vindiola et al. (8). A seed is deemed to be cooked when it can be squeezed easily. Boiling and steaming times, as well as pressure conditions, were selected from our previous paper (6). After cooking treatments, the legumes were drained and cooled to room temperature in covered plastic containers. Subsequently, cooked samples were frozen and then freeze-dried. Total Phenolic Quantification. Extraction of Polyphenols. The original raw legumes and the freeze-dried cooked legumes were ground to flour with an IKA all basic mill (IKA Works Inc., Wilmington, NC) to pass through a 60-mesh sieve. Extraction procedures were performed according to our earlier paper (9). The extracts of total phenolics were used for determination of total phenolics and evaluation of antioxidant activities. Determination of Total Phenolic Content (TPC). The TPC was determined by a Folin-Ciocalteu assay (10) with slight modifications (9) using gallic acid (GA) as the standard. The absorbance was
J. Agric. Food Chem., Vol. 57, No. 22, 2009
10719
measured using a UV-visible spectrophotometer (UV 160, Shimadzu, Kyoto, Japan) at 765 nm against a reagent blank. The TPC was expressed as milligrams of gallic acid equivalents per gram of dry legume (mg of GAE/g) through the calibration curve of gallic acid. The linearity range of the calibration curve was 50-1000 μg/mL (r = 0.99). Determination of Procyanidin Content (PAC). The PAC analysis was carried out according to the method of Broadhurst and Jones (11) and slightly modified in our laboratory (9). The absorption was measured using a UV-visible spectrophotometer (UV 160, Shimadzu) at 500 nm against methanol as a blank. The PAC was expressed as milligrams of catechin equivalents per gram of dry legume (mg of CAE/g) using the calibration curve of (þ)-catechin. The linearity range of the calibration curve was 50-1000 μg/mL (r = 0.99).
Quantification of Individual Free and Conjugated Phenolic Acid by HPLC. Extraction of Free Phenolic Acids. The extraction of free phenolic acids was performed by modifying the method of Luthria and Pastor-Corrales (12). Briefly, the raw and cooked legume samples (0.5 g in triplicate) were extracted twice, each with 5 mL of methanol/water/acetic acid/butylated hydroxytoluene (85:15:0.5:0.2, v/v/v/w) by shaking extraction tubes on an orbital shaker at 300 rpm at room temperature for 4 h. The extracts were concentrated at 45 °C under vacuum to remove solvents. The dry residue was dissolved in 5 mL of water and freeze-dried. The freeze-dried extracts (10 mg) were dissolved in 1 mL of 25% methanol. The methanol solution was centrifuged and then filtered through a 0.2 μm PVDF syringe filter and analyzed for free phenolic acid content by HPLC. Extraction of Conjugated Phenolic Acids. The extraction of conjugated phenolic acids was performed according to previous papers (12, 13) with slight modifications. Briefly, the raw and cooked legume samples (0.4 g in triplicate) were hydrolyzed and extracted with 10 mL of 2 N NaOH [containing 10 mM EDTA and 1% vitamin C (w/v)], at 4045 °C for 30 min. The reaction mixture was acidified by adding 2.8 mL of 7.2 N HCl. The mixture was vortexed for 5-10 s, and phenolic acids were extracted with ethyl acetate twice (2 10 mL). The combined organic layer was concentrated to dryness at 45 °C under vacuum to remove solvents. The dry residue was dissolved in 1.5 mL of 75% methanol. The methanol solution was filtered through a 0.2 μm PVDF syringe filter and analyzed for conjugated phenolic acid content by HPLC. HPLC Analysis of Phenolic Acids. The quantitative analysis of both free and conjugated phenolic acids was performed by HPLC according to our recent publication (14, 15). A Waters Associates (Milford, MA) chromatography system equipped with a model 720 system controller, a model 6000A solvent delivery system, a model 7125 loading sample injector, and a model 418 LC UV detector (270 nm) was used. A 4.6 mm 250 mm, 5 μm, Zorbax Stablebond Analytical SB-C18 column (Agilent Technologies, Rising Sun, MD) was used for separation at 40 °C, which was maintained with a column heater. All identified phenolic acids were quantified with external standards by using HPLC analysis as described previously (14). The phenolic acid contents were expressed as micrograms of phenolic acid per gram of legume (μg/g) on a dry weight basis.
Quantification of Flavan-3-ol and Flavonol by HPLC. Extraction of Flavonols. The cooked legume samples were freeze-dried and ground. The ground raw and cooked legumes (0.5 g in triplicate) were extracted according to the method described in our recent paper (15). HPLC Analysis of Flavonols. The quantitative analysis of flavonols was performed according to the methodology of isoflavone analysis developed by Murphy et al. (16) with a slight modification (15). The same Waters Associates chromatography system as used for phenolic acid analysis was used for the analysis of flavonols with 262 nm UV detection. A YMC-Pack ODS-AM-303 C18 reversed-phase column (250 mm 4.6 mm internal diameter, 5 μm particle size) was obtained from Waters and employed for chromatographic separation at 34 °C, which was maintained with a column heater.
Identification and Quantification of Flavan-3-ol and Flavonols. Five flavan-3-ols [(þ)-catechin, (þ)-epicatechin, epigallocatechin, epicatechin-gallate, epigallatecatechin-gallate (EGCG)] and nine flavonols or flavones (myricetin, luteolin, quercetin, apigenin, kaempferol, kaempferol-3-glucoside, kaempferol-3-rutinoside, quercetin-3-glucoside, quercetin-3-rutinoside) were commercially available and directly used to identify the sample peaks by comparing their retention times and HPLC profiles
10720
J. Agric. Food Chem., Vol. 57, No. 22, 2009
with those of the standard mixture. In addition, a spiking method was used for peak identification of some samples. External calibration curves were obtained for each of six external standards by plotting peak area of each standard against concentration. For the other flavonols without commercial standards, concentrations were calculated from the standard curves that were adjusted appropriately from the standard curves of the respective form of flavonols based on the differences in molecular weight and molar extinction coefficients of the compounds. Flavonol contents were expressed as micrograms of flavonol per gram of legume (μg/g) on a dry weight basis. Quantification of Anthocyanin by HPLC. The free phenolic acid extracts were also used for anthocyanin analysis, the analysis was performed on an HP 1090 series HPLC (Hewlett-Packard, Waldbronn, Germany) equipped with filter photometric detector, using a YMC Pack ODS-AM column (4.6 mm 250 mm, S-50 μm, 120A) according to our recent paper (14). The identifications and peak assignments of anthocyanins were primarily based on comparison of their retention times with those of the external standards and a blueberry reference sample. Standard curves of anthocyanins were plotted with peak areas against concentrations by duplicate injections of the six series of diluted working solutions of the standard mixture. Anthocyanin contents were expressed as micrograms of anthocyanin per gram of legume (μg/g) on a dry weight basis.
Extraction and Determination of Total Saponin. Extraction of Saponin. Extraction procedures were performed by modifying the method of Makkar and Becker (17). Briefly, the raw and freeze-dried cooked legume flours (0.5 g in triplicate) were defatted with 10 mL of petroleum ether by shaking for 4 h, and then the residues were extracted twice, each with 5 mL of 80% aqueous methanol, on an orbit shaker by shaking for 4 h each time. The extracts were stored at 4 °C in the dark for use. Determination of Total Saponin Content (TSC). The TSC was determined using the spectrophotometric method described by Hiai et al. (18). Briefly, 0.1 mL of legume extract, 0.4 mL of 80% methanol solution, 0.5 mL of freshly prepared 8% vanillin solution (in ethanol), and 5.0 mL of 72% sulfuric acid were mixed well in an ice-water bath. The mixture was warmed in a water bath at 60 °C for 10 min and then cooled in ice-cold water. Absorbance at 544 nm was recorded against the reagent blank with a UV-visible spectrophotometer (UV 160, Shimadzu). The results were expressed as milligrams of soyasaponin equivalent per gram of legume (mg of SSE/g) on a dry weight basis from a standard curve of different concentrations of crude soyasaponin (contained a minimum of 80% saponin, Sigma-Aldrich) in 80% aqueous methanol.
Extraction and Determination of Phytic Acid. Extraction of Phytic Acid. Phytic acid in the legume was extracted according to the method of Gao et al. (19). Briefly, the raw and freeze-dried cooked legume flours (0.5 g in triplicate) were defatted with 10 mL of petroleum ether by shaking on an orbit shaker for 4 h, and then the residues were extracted with 10 mL of 2.4% HCl by shaking on the orbit shaker for 16 h. The extracts were stored at 4 °C in the dark for further analysis. Determination of Phytic Acid. The phytic acid was determined using the colorimetric (Wade Reagent) method described by Gao et al. (19) with slight modification. Briefly, 0.1 mL of legume extract was diluted by 2.9 mL of distilled water, and then 3 mL of this diluted sample was combined with 1 mL of freshly prepared Wade reagent (0.03% FeCl3 3 6H2O þ 0.3% sulfosalicylic acid) in a 15 mL VWR tube. The contents were thoroughly mixed on a vortex and centrifuged at 5500 rpm at 10 °C for 10 min. A series of calibration standards containing 0, 5, 10, 20, 25, 50, 75, or 100 μg/mL of phytic acid were prepared by diluting 10 mg/ mL of phytic acid stock solution with distilled water. Absorbance of color reaction products for both samples and standards was read at 500 nm on a UV-visible spectrophotometer (UV 160, Shimadzu) against water as blank. The results were expressed as milligrams of phytic acid per gram of legume (mg of PA/g) on a dry weight basis.
Determination of Chemical Antioxidant Capacities. Determination of Ferric Reducing Antioxidant Power (FRAP). The FRAP assay was performed according to the method described by Benzie and Strain (20). The total phenolic extract was first properly diluted with deionized water to fit within the linearity range. The absorbance was measured using a UV-visible spectrophotometer (UV 160, Shimadzu) at 593 nm against reagent blank. The FRAP value was expressed as
Xu and Chang 2þ
millimoles of Fe equivalents per 100 g of dry legumes (mmol of FE/ 100 g) using the calibration curve of Fe2þ. The linearity range of the calibration curve was 0.1-1.0 mM (r = 0.99).
Determination of Peroxyl Radical Scavenging Capacity (PRSC). The PRSC assay was performed in a cell-free system according to the method validated by Adom and Liu (21) with modifications. Briefly, 1 mM DCFH-DA solution was obtained from a 20 mM stock solution (DCFH-DA dissolved in methanol) by dilution with PBS buffer. Just prior to use, an aliquot (400 μL) of 1 mM DCFH-DA solution was added into 3.6 mL of 1.0 mM KOH and hydrolyzed for 3-5 min to remove the diacetate (DA) moiety, and then DCFH was diluted to 10 μM (as final working solution) with prewarmed PBS at 37 °C. Twenty microliters of suitably diluted legume extracts, blank, and Trolox calibration solutions were loaded into clear 96-well microplates in triplicate based on a balanced layout. AAPH was used as the peroxyl generator; 27.2 mg of AAPH solid was dissolved (just before PRSC measuring) into 5 mL of warmed HBSS. Immediately after dissolving, 20 μL of AAPH solution was dispensed through a pump and an autoinjector of the plate reader into appropriate wells according to a balanced layout. The BMG Fluostar Optima Microplate Reader (BMG Labtech GmbH, Offenburg, Germany) was programmed to record the fluorescence of dichlorofluorescein (DCF) on each cycle. Kinetic readings were measured with emission at 520 nm and excitation at 485 nm for 1 h with 85 s per cycle setting. The kinetics of the fluorescence were recorded by the software BMG OPTIMA running on a PC. The areas under the average fluorescence-reaction time kinetic curve (AUC) for both control and samples were integrated and used as the basis for calculating PRSC antioxidant activity. The net AUC was obtained by subtracting the AUC of the blank from that of a sample or standard, expressed as net AUC = AUCsample - AUCblank. The quantification method is similar to the ORAC assay in our previously published paper (11). PRSC values were expressed as micromoles of Trolox equivalent per gram legume (μmol of TE/g) on a dry weight basis. Cellular Antioxidant Activity (CAA) Assay. Human gastric adenocarcinoma AGS cells were grown in complete growth medium F-12K (Mediatech, Inc.) supplemented with 10% FBS and 1% penicillinstreptomycin (v/v). Cells were maintained in a humidified 5% CO2 incubator at 37 °C. Cells used in this study were between passages 47 and 51. The CAA assay was performed by modifying the methods of Eberhardt et al. (22) and Wolfe and Liu (23) using a rapid proliferating gastric adenocarcinoma cell line AGS. Briefly, AGS cells were seeded at a density of 6 104/well on a 96-well microplate in 100 μL of complete growth medium. The outside wells of the plate were filled with 200 μL of PBS to maintain the temperature and prevent medium evaporation of the inner wells. After 24 h of culturing, medium was removed and wells were washed with prewarmed PBS twice. Attached AGS cells were treated with 20 μL of various concentrations of legume extracts, and 180 μL of prewarmed treatment medium (EMEM, phenol free, FBS free) contained final 25 μM DCFH-DA for 1 h. Subsequently, treatment medium was removed, and wells were washed twice with 150 μL of PBS to remove medium, extracellular sample residue, and fluorescence substance. Then 80 μL of HBSS (prewarmed at 37 °C in a water bath) was added to wells, and the microplate was incubated in the BMG Fluostar Optima Microplate Reader (BMG Labtech GmbH, Offenburg, Germany) for a minimum of 10 min to maintain the temperature evenly for each well at 37 °C. Just prior to the assay, 25 mg of AAPH dry solid was dissolved into 5 mL of prewarmed HBSS in a 15 mL of tube. Immediately after dissolving, 20 μL of AAPH solution was dispensed through a pump and an autoinjector into appropriate wells according to a balanced layout. The BMG plate reader was programmed to record the fluorescence of DCF on each cycle. Kinetic readings were measured with emission at 520 nm and excitation at 485 nm for 1 h with 85 s per cycle setting. Each plate included at least five blank and five control wells. The blank wells contained cells treated with DCFH-DA and HBSS without oxidant AAPH and antioxidant samples. The control wells contained cells treated with DCFH-DA, HBSS, and oxidant AAPH without antioxidant samples. Quantification of CAA. The data were analyzed using Microsoft Excel (Microsoft, Roselle, IL). The area under the curve (AUC) was calculated as AUC = [R1/2 þ sum (R2:Rn-1) þ Rn/2] CT, where R1 is the fluorescence reading at the initiation of the reaction, Rn is the last measurement, and CT = cycle time in minutes. The net AUC was obtained by subtracting the AUC of the blank from that of a sample or
Article standard, expressed as net AUC = AUCsample - AUCblank. The CAA unit was expressed as CAA unit = 100 - (net AUCsample/net AUCcontrol) 100. The median effective concentration (EC50) was defined as the dose required to cause a 50% inhibition for sample extract or standard compound and calculated through the software CurveExpert (version 1.3). Antiproliferation Assay. Cell Lines and Cell Cultivation. Three typical human cancer cell lines were chosen for antiproliferation assays due to their rapid proliferation properties and easy maintenance: (1) Gastric adenocarcinoma cell AGS was maintained in F-12K medium. (2) Colorectal adenocarcinoma cell SW480 was maintained in L-15 medium. (3) Prostate carcinoma cell DU145 was maintained in EMEM. All media were supplemented with 10% FBS and 1% penicillin-streptomycin. Cells were maintained at 37 °C and 5% CO2 (except for cell line SW480 without CO2). Cell culture medium and cultivation conditions were chosen as above according to the suggestion of ATCC. Routine observation for cell viability was performed under phase contrast inverted microscopy. Cell numbers were determined by trypan blue exclusion method and counting in a hemocytometer. MTT Assay. The antiproliferation assay should be performed under solvent-free conditions to eliminate solvent effects. Therefore, a portion of hydrophilic total phenolic extract was freeze-dried, and then the freezedried extract (10 mg) was dissolved in cell culture medium as stock sample solution. The final concentrations of samples (0.125, 0.25, 0.5, 1, 2, and 5 mg/mL) were obtained by diluting sample with medium. The antiproliferation assay was performed according to a well-established MTT method (24). Briefly, exponentially growing cells were seeded into 96-well culture plates at a seeding density of 1 104 cells/well in 180 μL of medium. After 24 h, attached cells were exposed to the legume extracts with final concentrations as above for 48 h. Subsequently, 20 μL of MTT (5 mg/mL) was added to each well. The microplate was placed back on the incubator and cultured for an additional 4 h. Subsequently, the culture media were sucked from the wells, leaving cells that were adhered to the plates. Then 150 μL of DMSO was added into each well to dissolve yellow formazan (product of the reduction of tetrazolium by viable cells), and then the microplate was gently shaken on an orbit shaker for 10-15 min in the dark. The reaction resulted in the reduction of MTT by the mitochondrial dehydrogenases of viable cells to a purple formazan product, which was measured at 540 nm by the BMG microplate reader. The 50% growth inhibitory concentration (IC50) was defined as the legume concentration (units in mg/mL) required to cause a 50% inhibition and was used as the basis for comparing antiproliferation activity of different samples. Statistical Analysis. All boiling and steaming processes were performed in triplicate. Further composition analyses and antioxidant evaluations were performed on the basis of triplicate processed samples. The data were expressed as mean ( standard deviation. Statistical analysis was performed using 2005 SAS (version 9.1, SAS Institute Inc., Cary, NC). Duncan’s multiple-range tests were used to determine the significant differences at p < 0.05. RESULTS AND DISCUSSION
Effect of Thermal Processing on Total Phenolics and Procyanidins of CSFL’s. TPC and PAC of the raw and cooked CSFLs are presented in Figure 1. Significant differences (p < 0.05) in TPC (Figure 1A) and PAC (Figure 1B) were found among most processing treatments of green pea, yellow pea, chickpea, and lentil. No significant differences in TPC existed between the two boiling treatments (regular and pressure) for green, yellow pea, and lentil, but significant differences (p < 0.05) existed between conventional and pressure boiling of chickpea. Significant differences (p < 0.05) in TPC existed between conventional and pressure steaming treatments of all tested CSFLs. Significant differences in PAC existed between conventional and pressure boiling treatments as well as between conventional steaming and pressure steaming. The TPC and PAC of cooked CSFLs were significantly reduced as compared to the respective original uncooked CSFLs. Approximately 40-60% of TPC in green pea, yellow pea, and chickpea and 60% of TPC in lentil were reduced (Table 1) after boiling processing, whereas about 10-30% of TPC in green pea, yellow pea, and chickpea and
J. Agric. Food Chem., Vol. 57, No. 22, 2009
10721
50-60% of TPC in lentil were lost by steaming. Pressure processing (both boiling and steaming) lost relatively less TPC than regular processing for green pea, yellow pea, and chickpea due to shorter processing times. However, pressure steaming lost more TPC than conventional steaming in the case of lentil. Effect of Thermal Processing on Saponins and Phytic Acids of CSFLs. Figure 2 shows the saponin and phytic acid contents in raw and cooked CSFLs. The level of saponin (soyasaponin equivalent) in raw CSFLs ranged from 9.82 to 17.78 mg/g (Figure 2A). The level of phytic acid in raw CSFLs ranged from 7.51 to 18.92 mg/g (Figure 2B). Among the raw legumes, lentil and chickpea possessed higher saponin content than green pea and yellow pea. On the other hand, green pea and chickpea had higher phytic acid content than yellow pea and lentil. As compared to the raw CSFLs, all cooking treatments significantly (p < 0.05) reduced both saponin and phytic acid contents. Different cooking methods studied have varied effects in reducing the level of saponin and phytic acids. Among the cooking treatments, boiling appeared to effectively reduce the saponin and phytic acid levels in all CSFLs. The reduction ranges (Table 1) on cooking were 5.2-42.9% for saponin and 15.0-24.3% for phytic acid. The decrease in phytic acid content of lentil after cooking was 21.6-21.9%. These results are comparable to the study of Wang et al. (25), who found cooking caused a 15.9% reduction in phytate levels, but in contrast to those of Elhardallou and Walker (26), who found cooking lentils caused a 60.5% reduction. The discrepancies may be due to the differences in sample sources or processing methods. The apparent decrease in phytic acid content during cooking may be partly due either to the formation of insoluble complexes between phytic acid and other components, such as phytate-protein and phytate-proteinmineral complexes, or to the inositol hexaphosphate hydrolyzed to penta- and tetraphosphates (27). Traditionally, saponins and phytic acids have been considered to be antinutritional factors. The presence of these antinutritional components in legumes impairs the digestion of protein, decreases Ca, Fe, and Zn bioavailability, and therefore reduces the nutritional value of legumes. However, recent evidence indicates that low levels of phytic acid had healthful effects as antioxidant (28). Reductions in glycemic response to starchy foods as well as lower plasma cholesterol and triglyceride levels have been observed with endogenous phytate consumed in foods or with the addition of purified sodium phytate. Therefore, phytate may play an important role in controlling hypercholesterolemia and atherosclerosis. In addition, phytic acid had shown anticancer effects in the colon and mammary gland in rodent models and in various tumor cell lines in vitro (29). Therefore, reduction of phytic acid is expected to enhance the bioavailability of proteins and dietary minerals of legumes, and at the same time the lower level of phytic acid may still have some health promotional activities. In view of these beneficial effects, the term “antinutrient” used to describe food constituents such as phytic acid needs to be re-evaluated (30). Effect of Thermal Processing on Antioxidant Capacities of CSFLs. Antioxidant activity determination is reaction mechanismdependent. The specificity and sensitivity of a single method do not lead to the complete examination of all phytochemicals in the extract. Therefore, a combination of several tests could provide a more reliable assessment of the antioxidant activity profiles of legume samples. Previously, boiling and steaming effects on antioxidant activities in terms of DPPH radical scavenging activity and oxygen radical absorbing capacity (ORAC) of CSFLs have been reported in our earlier paper (6). However, thermal processing effects on the ferric reducing antioxidant power (FRAP), peroxyl radical scavenging capacities
10722
J. Agric. Food Chem., Vol. 57, No. 22, 2009
Xu and Chang
Figure 1. Effect of thermal processing on phenolics (μg/g) of cool-season food legumes. Bar data are expressed as mean ( standard deviation (n = 3) on a dry weight basis. The same letter above the bar indicates no significant difference (p < 0.05) within each group of legumes.
(PRSC), and cell-based antioxidant capacities of CSFLs have not been documented. Therefore, the current study was performed to investigate these activities of cooked CSFLs on the basis of four processing conditions selected from our previous study. The chemical antioxidant activities (FRAP and PRSC) and cellular antioxidant activities (CAA) of the raw and cooked CSFLs are presented in Figure 3 and Table 2, respectively. Significant differences (p
